It seems that when you look at the way the data was gathered, and find that while some areas of the GBR declined, others gained. Analysed other ways, it doesn’t seem so alarming.
Perhaps the phrase is “coral picking”?
Assessing loss of coral cover on the Great Barrier Reef: A response to Hughes et al. (2011)
Hughes et al. (Coral Reefs, 2011, in press) challenge our interpretations of the changes in coral cover observed on the Great Barrier Reef (GBR) between 1986 and 2004 (Sweatman et al. in Coral Reefs 30:521–531, 2011). They question whether we can accurately assign all causes of coral loss; we contend that this makes no difference to the observed changes. They defend the validity of historical data on coral cover from before the start of systematic large-scale monitoring and conclude that coral cover has been declining since at least 1960, but we find no trend in the early data. We remain convinced that combining data collected at different spatial scales (quadrats and transects in the past mixed with more recent whole-reef averages from manta tows) are likely to overestimate decline, because whole-reef averages will very rarely reach the high cover values that can occur at the quadrat scale.
Hughes et al. (Coral Reefs, 2011, in press) state that we dismiss runoff as a cause of ecosystem degradation; we defend our interpretations and dispute some of their examples. In summary, we stand by our conclusion that coral cover on the GBR declined in the period 1986–2004 but through localised and unsynchronised changes that included recovery.
Click image below for preview.